S TATISTICAL

PROCESS CONTROL
Pradeep Kumar
Professor
Mechanical & Industrial Engineering
Indian Institute of Technology Roorkee
 Modern Importance of Quality

“The first job we have, is to turn out quality

products that consumers will buy and keep on
buying. If we produce it efficiently and
economically, we will earn a profit, in which you
will share.”
- William Cooper Procter
Grandson of the founder of Procter and Gamble
(October 1887)

3
 Three Issues

1. Productivity
2. Cost
3. Quality

4
 Significant Issue
Off These three determinants of the profitability the most
significant factor in determining the long range success or failure
of an organization is Quality.

Good quality of product and service can provide:

1. Competitive edge

2. Reduce cost due to returns, rework, and scrap

3. Productivity and Profits
4. Generates Satisfied customers:
Continued Patronage and word of mouth advertisement

5
 Cost of Quality

Running a company by profit alone is

like driving a car by looking in the
rearview mirror.
It tells you where you’ve been, not where
you are going!
Dr. E. Deming
 What is the Cost of Quality?
• Cost of Quality: the cost of ensuring that the
job is done right + the cost of not doing the job
right.

Cost of Conformance + Cost of Non-Conformance

(Prevention & Appraisal) (Internal/External Defects)

7
 Four Cost Categories Related to Quality

•• Prevention
Preventioncost:
cost:Cost
Costof
ofplanning
planningand
andexecuting
executingaaproject
projectso
soititisis
error-free
error-freeor
orwithin
withinan
anacceptable
acceptableerror
errorrange.
range.
•• Appraisal
Appraisalcost:
cost:Cost
Costof
ofevaluating
evaluatingprocesses
processesand
andtheir
theiroutputs
outputsto
to
ensure
ensurequality.
quality.
•• Internal
Internalfailure
failurecost:
cost:Cost
Costincurred
incurredto
tocorrect
correctan
anidentified
identifieddefect
defect
before
beforethe
thecustomer
customerreceives
receivesthe
theproduct.
product.
•• External
Externalfailure
failurecost:
cost:Cost
Costthat
thatrelates
relatesto
toall
allerrors
errorsnot
notdetected
detected
and
andcorrected
correctedbefore
beforedelivery
deliveryto
tothe
thecustomer.
customer.

8
Normal Cost of Quality Distribution when
Quality System is NOT in Place

Appraisal Cost

Prevention
Cost

Internal
Failure Cost
Cost of
Quality
External
Failure
Cost

9
Optimum Cost of Quality Distribution
when Quality System is in Place

Appraisal Cost

Prevention
Cost of Cost
Quality Internal Failure Cost

External Failure Cost

10
 Definitions and Meaning of Quality

The Eight Dimensions of Quality

1. Performance (will the product do the intended job?)

2. Reliability (how often the product fails?)
3. Durability (how long the product lasts?)
4. Serviceability (how easy is to repair the product?)
5. Aesthetics (what does the product look like?)
6. Features (what does the product do?)
7. Conformance to (Whether the product meet the
Standards specifications)
8. Perceived quality (what is the reputation of a company or its
products?)

11
12
 Have you ever…

• Shot a rifle?
• Played darts?
• Played basketball?

What is System of point of

these sports?
13
 Have you ever…

Shot a rifle?
Played darts?
Played basketball? Ram

Arjun

Who is the better shot? 14

 Discussion

• What do you measure in your process?

• Why do those measures matter?
• Are those measures consistently the same?
• Why not?

15
 Variability

• Deviation = distance between 8

7
observations and the mean (or 10
average) 8
9

Ram
Observations Deviations
10 10 - 8.4 = 1.6
9 9 – 8.4 = 0.6
8 8 – 8.4 = -0.4
8 8 – 8.4 = -0.4
7 7 – 8.4 = -1.4
averages 8.4 0.0
Arjun

16
 Variability

• Deviation = distance between

observations and the mean (or
average) Ram

Observations Deviations
7
7 7 – 6.6 = 0.4 6
7 7 – 6.6 = 0.4 7
7 7 – 6.6 = 0.4 7
6 Arjun
6 6 – 6.6 = -0.6
6 6 – 6.6 = -0.6
averages 6.6 0.0

17
 Variability
8
7
10
• Variance = average distance between 8
observations and the mean squared 9

Ram

Observations Deviations Squared Deviations

10 10 - 8.4 = 1.6 2.56
9 9 – 8.4 = 0.6 0.36
8 8 – 8.4 = -0.4 0.16
Arjun
8 8 – 8.4 = -0.4 0.16
7 7 – 8.4 = -1.4 1.96
averages 8.4 0.0 1.0 Variance
18
 Variability

• Variance = average distance between

observations and the mean squared
Ram

Observations Deviations Squared Deviations

7
7 7 - 6.6 = 0.4 0.16
6
7 7 - 6.6 = 0.4 0.16 7
7 7 - 6.6 = 0.4 0.16 7
6 6 – 6.6 = -0.6 0.36 6 Arjun
6 6 – 6.6 = -0.6 0.36
averages 6.6 0.0 0.24 Variance
19
 Variability

• Standard deviation = square root

of variance
Ram

Variance Standard
Deviation
Ram 1.0 1.0
Arjun 0.24 0.4898979 Arjun

But what good is a standard deviation

20
 Variability

The world tends to

be bell-shaped

Even very rare Fewer Most Fewer Even very rare

outcomes are in the outcomes in the outcomes are
possible “tails” occur in the “tails” possible
(probability > 0) (lower) middle (upper) (probability > 0)

21
 Variability

Here is why: Even outcomes that are equally

likely (like dice), when you add
them up, become bell shaped

Add up the dots on the dice

0.2

0.15
Probability

1 die
0.1 2 dice

0.05 3 dice

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Sum of dots
22
 “Normal” bell shaped curve

Add up about 30 or most things

and you start to be “normal”

Normal distributions are divided up

into 3 standard deviations on
each side of the mean

Once you know that, you

know a lot about
what is going on

And that is what a standard deviation

is good for 23
 Potential Reasons of Variations
The Operator
• Training, supervision, technique

The Method
• Procedure, Set-up, temperature, etc.

The Material
• Moisture content, blending, contamination

The Machine
• Machine condition, inherent precision
Management
• Poor Process management, Poor System

24
 Causes of Variations
Common Causes : Improvement in system
• Variation inherent in process
• Can be eliminated only through improvement in system
• No pattern

Assignable Causes : Control of Process

• Variation due to identifiable factors
• Can be modified through operator or management action
• May exhibit a pattern
• Examples of special causes include: wrong tool, wrong production
method, improper raw material, operator’s skill, wrong die etc.

SPC uses samples to identify that Assignable causes have occurred25

 Statistical Process Control
• The underlying concept of statistical process control is based on a
comparison of what is happening today with what happened previously.

• We take a snapshot of how the process typically performs or build a

model of how we think the process will perform and calculate control
limits for the expected measurements of the output of the process.

• Then we collect data from the process and compare the data to the control
limits.

The majority of measurements should fall within the control limits.

Measurements that fall outside the control limits are examined to see
if they belong to the same population as our initial snapshot or model.

26
 Statistical Quality Control
• The purpose of statistical quality control is to ensure, in a cost efficient
manner, that the product shipped to customers meets their specifications.

• Inspecting every product is costly and inefficient, but the

consequences of shipping non conforming product can be significant
in terms of customer dissatisfaction.

• Statistical Quality Control is the process of inspecting enough

product from given lots to ensure a specified quality level.

27
 What are SPC Techniques?

There are many ways to implement process control.

Key monitoring and investigating tools include:
• Histograms
• Run Charts
• Pareto Charts
• Cause and Effect Diagrams
• Flow Diagram
• Scatter Diagrams
• Control Charts

28
 Control Charts:
Recognizing Sources of Variation
• Why Use a Control Chart?
– To monitor, control, and improve process performance over time by
studying variation and its source.

What Does a Control Chart Do?

• Focuses attention on detecting and monitoring process variation over
time;
• Distinguishes special from common causes of variation, as a guide to
local or management action;
• Serves as a tool for ongoing control of a process;
• Helps improve a process to perform consistently for higher quality,
lower cost, and higher effective capacity;
• Provides a common language for discussing process performance. 29
 Control Charts:
Recognizing Sources of Variation
How Do I Use Control Charts?
 There are many types of control charts. The control
charts that you or your team decides to use should be
determined by the type of data that you have.

 Data are of two types

– Variables
– Attributes

30
 Control Charts:
Recognizing Sources of Variation

Variables “Things we measure”

Length Blood Pressure Diameter

Weight Volume Tensile strength
Time Temperature

Attributes “Things we count”

Number or percent defective items

Number of defects per item
Types of defects etc.

31
 Control Charts: Types
Control Charts for Variables Data

_
X and R charts : for sample averages and ranges
_
X and s charts : for sample averages and standard deviations

Md and R charts : for sample medians and ranges

X and Rm charts : for individual measures and moving ranges

Control Charts for Attributes Data

p charts : proportions of units nonconforming

np charts : number of units nonconforming

c charts : number of defects/nonconformities

u charts : number of defects/nonconformities per unit 32

 Control Charts: Selection

• How Do I Select Control Charts?

– Use the following tree diagram to determine which

chart will best fit your situation.
– Only the most common types of charts are addressed.

33
Control Chart Selection: Variable Data

n=1 2<n<9 n is ‘small’ n is ‘large’

median 3<n<5 n > 10

X & Rm X&R X&R X&S

34
Control Chart Selection: Attribute Data

Defect or
Defective Data
Nonconformity Data

Constant Variable Constant Variable

sample size sample size n > 50 n > 50

c chart u chart p or np chart p chart

35
 Basis of Control Chart
C

L
=
E
θ

θ
E
s
t
i
m
a
t
e
s
o
f
θ


   
U
C
L
=
E
θ
+
K
σ
θ
S
.
D
o
f
E
S
T
I
M
A
T
O
R
θ


 
L
C
L
=
E
θ
-
K
σ
θ
K: - No. of std. deviations of the sample statistics that the
control limits are placed from the centerline.

If we assume normal distribution k=3, 99.74% fall in the range of UCL & LCL
0.26% or 0.0026 fraction out of range.
K can be decided on % outside the limits. 0.2%--K=3.09

36
 Charts may signal incorrectly!

Charts repeatedly apply hypothesis testing!

– Type I error with charts:

Concluding that a process is not in control
when it actually is
– Type II error with charts:
Concluding that a process is in control when
it is not
37
 Two types of error
Type I error: - results from inferring that a process is out of control
when it is actually in control.

Prob. of type I error is 

Type I error = 

= sum of two tails area outside
control limits

38
 Two types of error
Type II error: - results from inferring that a process is in
control when it is actually out of control.
(no of observations goes out of control limit)

39
 Example
A control chart is to be constructed for the average
breaking strength of nylon fibers. Samples of size 5 are randomly
chosen from the process. The process mean and standard
deviation are estimated to 120 kg and 8 kg, respectively.
(a). If the control limits are placed 3 standard deviations from
the process mean, what is the probability of a Type I error?
Solution: From the problem statement,  = 120 and s = 8. The
center line for the control chart is at 120 kg. The control limits
are
Mean ± 3 sd of process
sd of process = sd of process mean / square root of sample size

40
Example

41
 Example

Since the control limits are 3 standard deviations from the

mean, the standardized normal value at the upper control limit is

X  
Z 
X
1 3 0 .7 3 3  1 2 0
  3 .0 0
8 / 5

Similarly, the Z-value at the lower control limit is - 3.00. For

these Z-values is the standard normal table in Appendix A-3, each
tail area is from to be.0013.The probability of a Type I error, as
shown by the shaded tail area is = 0.0026
42
 Example

(b). If the process mean shifts to 125kg, what is the

probability of concluding that the process is in control and hence
making a Type II error on the first sample plotted after the shift?

130.733-125 109.267 125

Z1 = =1.60 Z2   4.40
8/ 5 8/ 5

From the standard normal table in Appendix A-3, the tail

area above the upper control limit is .0548. The standardized
normal value at the lower control limit is From Appendix A-3,
the tail area below the lower control limit is approximately .0000.
The area between the control limits is 1 - (.0548 + 0000) = .9452.
43
 Operating Characteristic Curve

An operating characteristic (OC) curve is a measure of

goodness of a control chart's ability to detect changes in process
parameters.

Specifically, it is a plot of the probability of the Type II

error versus the shifting of a process parameter value from
its in-control value.

OC curves enable us to determine the chances of not

detecting a shift of a certain magnitude in a process parameter on
a control chart.
44
Operating Characteristic Curve

45
 Operating Characteristic Curve
Refer to the data in Example earlier involving the control
chart for average breaking strength of nylon fibers. Samples of
size 5 are randomly chosen from a process whose mean and
standard deviation are estimated to be 120 kg and 8kg,
respectively. Construct the operating characteristic curve for
increases in the process mean from 120 kg.

Z-value at Z-value at Probability of

Process Area above Area above
UCL LCL nondetection,
mean UCL LCL
Z1 Z2 
123.578 2.00 .0228 -4.00 .0000 .9772
127.156 1.00 .1587 -5.00 .0000 .8413
130.733 .00 .5000 -6.00 .0000 .5000
134.311 -1.00 .8413 -7.00 .0000 .1587
137.888 -2.00 .9772 -8.00 .0000 .0228
141.466 -3.00 .9987 -9.00 .0000 .0013
46
Operating Characteristic Curve

47
 Control Chart Performance
Average Run Length
• The average run length (ARL) is a measure of the
performance of the control chart. This denotes the
number of samples on average, required to detect an out-
of-control signal.
• Let p = probability that any point exceeds the control
limits. Then,

For a process in control ARL = 1/α = 1/0.0027 = 371 ( False Alarm)

For a process out-of-control control ARL = 1/(1-β) = 1/(1-0.9192) = 12

48
 Control Chart Performance
Magnitude of Process Shift α Error β Error ARL Remarks
0 0.0026 - 370.4 Process in Control
0.5 σ - 0.9936 155.2 Process Out of Control
1.0 σ - 0.9772 43.9 Process Out of Control
1.5 σ - 0.9334 15 Process Out of Control
2.0 σ - 0.8412 6.3 Process Out of Control
3.0 σ - 0.5000 2 Process Out of Control

For a process in control we For an out of control process it is

Prefer the ARL to be large desirable for the ARL to be small
Because an observation we want to detect the out of control
Plotting outside the control condition.
Limit represent false alarm.

49
Effect of CL on errors in inference making

Type I error (α) reduces when control limits are placed farther apart.
Control Limits > = 4 sigma , Type I error is negligible.
Type II error (β) α [1/ α]

Type II error increases when Type I error decreases.

Type II error α [1/ n].

μ
±
k
σ

Control limits =  

 
 ± σx / n
μ
50
Effect of sample size on errors in inference making

n1<n2<n3

51
 Warning Limits on Control Charts

–Warning limits (if used) are typically set at 2 standard deviations from the
mean.
–If one or more points fall between the warning limits and the control limits, or
close to the warning limits the process may not be operating properly.
–Good thing: Warning limits often increase the sensitivity of the control chart.
–Bad thing: Warning limits could result in an increased risk of false alarms.

52
 General Rules for Out of Control Situations

Natural Pattern:
Np are indicative of a process that is in control; i.e. they
demonstrate the presence of a stable system of common cause.

Sudden Shift:
Sudden change or jump occurs because of changes – in process
setting as temp, pr., depth of cut, change in customer behavior no.
of tellers changed, new eq, process etc.

53
 General Rules for Out of Control Situations
Gradual Shift :

- Change in the Q. of coming material overtime 

X  
 - Change in the maintenance program 
 new operator 
 
R  decrease in worker skill 
 gradual improvement in vendors supply 
 

54
 General Rules for Out of Control Situations

Trending Pattern:

X  Tool wear, die wear, gradual deterioration of Eq.

R  Operator skill 

55
 General Rules for Out of Control Situations

Cyclic Pattern:

 repetitive periodic behavior

X : Periodic change in temp rotation of operator, seasonal variation
R : Operators fatigue periodic maintenance

56
 Control Chart Construction
• Select the process to be charted;
• Decide what to measure and count; (Pareto sizing)
• Determine sampling method and plan;

– How large a sample needs to be selected? Balance the time and cost to collect a
sample with the amount of information you will gather.
– As much as possible, obtain the samples under the same technical conditions: the
same machine, operator, lot, and so on. (Rational Sub-grouping)
– Frequency of sampling will depend on whether you are able to discern patterns in
the data. Consider hourly, daily, shifts, monthly, annually, lots, and so on. Once
the process is “in control”, you might consider reducing the frequency with which
you sample.
– Generally, collect 20-25 groups of samples before calculating the statistics and
control limits.

57
 Control Chart Construction

• Initiate data collection:

– Run the process untouched, and gather sampled data.
– Record data on an appropriate Control Chart sheet or
other graph paper. Include any unusual events that occur.
• Calculate the appropriate statistics and control limits:
– Use the appropriate formulas.
• Construct the control chart(s) and plot the data.

58
 Control Chart Interpretation

What is Process Control?

Process Control is the active changing of the process

based on the results of process monitoring.

Once the process monitoring tools have detected an out-

of-control situation,

the person responsible for the process makes a change to

bring the process back into control.

59
 Control Chart Interpretation

What to do if the process is "Out of Control"?

If the process is out-of-control, the process engineer looks

for an assignable cause by following the out-of-control
action plan (OCAP) associated with the control chart.

60
 Set of Control Chart Interpretation
rules to determine "Out of Control"

General rules for detecting out of control or non-random situations

Any Point Above +3 Sigma
------------------------------------------------------------------------ +3 LIMIT
2 Out of the Last 3 Points Above +2 Sigma
---------------------------------------------------------------------- +2 LIMIT
4 Out of the Last 5 Points Above +1 Sigma
---------------------------------------------------------------------- +1 LIMIT
8 Consecutive Points on This Side of Control Line
=========================================== CENTER LINE
8 Consecutive Points on This Side of Control Line
------------------------------------------------------------------------ -1 LIMIT
4 Out of the Last 5 Points Below - 1 Sigma
-------------------------------------------------------------------------- -2 LIMIT
2 Out of the Last 3 Points Below -2 Sigma
------------------------------------------------------------------------- -3 LIMIT
Any Point Below -3 Sigma

TREND RULE :6 in a row trending up or down.

14 in a row alternating up and down
61
 Control Chart Representing Limits, Special
Causes, Common Causes

Special cause Common cause

Upper
control limits

Centre Line

Lower
control limits

Stable process Unstable process

62
Control Charts for Variable Type of Data

and R charts

In the x bar chart the sample means are plotted in order to

control the mean value of a variable.

In R chart, the sample ranges are plotted in order to control

the variability of a variable

63
 Control Charts for Variable Type of Data
Centre line, upper, & lower control limit for R charts are calculated.
The formulae used are as following:

Ri  X max ( i )
 X min( i )

Ri = range of ith sample

Xmax(i) = maximum value of the data in ith sample
Xmin (i) = minimum value of the data in ith sample

 R i
= mean of ‘g’ samples
R  i1 R
g
Where D3 and D4 are
UCLR  D4 R (Upper control limit for R chart) constants and their
values can be obtained
from standard tables.
LCLR  D3 R (Lower control limit for R chart) These values depend
upon the size of sample.
64
Control Charts for Variable Type of Data
Factors for Calculating Limits for and R Charts

n A2 D3 D4
2 1.880 0 3.267
3 1.023 0 2.575
4 0.729 0 2.282
5 0.577 0 2.115
6 0.483 0 2.004
7 0.419 0.076 1.924
8 0.373 0.136 1.864
9 0.337 0.184 1.816
10 0.308 0.223 1.777

65
 Control Charts for Variable Type of Data
Centre line, upper, & lower control limit for x bar chart are calculated.
The formulae used are as following:
n

X i
Xi  i 1

X i = mean of the ith sample

n = sample size,
Xi = ith data
g

X
X  i 1
i
 CL X = mean of ‘g’ samples
g x
Where A2 is
constants and its
value can be
UCL x  X  A2 R (Upper control limit for X bar chart) obtained from
standard table. Its
LCL x  X  A2 R (Lower control limit for X bar chart) value depend
upon the size of
sample. 66
 Control Charts for Variable Type of Data
Example
Mean values and ranges of data from 20 samples (sample size = 4) are shown
in the table below:

Mean Mean Mean Mean Mean

S.N of Range S.N of Range S.N of Range S.N of Range S.N of Range
Sample Sample Sample Sample Sample

1 10 4 5 9 5 9 10 4 13 12 4 17 12 4

2 15 4 6 11 6 10 11 6 14 12 3 18 15 3

3 12 5 7 11 4 11 12 5 15 11 3 19 11 3

4 11 4 8 9 4 12 13 4 16 15 4 20 10 4

20

 R
Average of Ranges of 20 samples = i  1
20
= 4.15 (Center Line of R Chart)
Upper Control Limit of R chart = D4*4.15 (D4= 2.282 for sample size 4)
= 9.47  9.5
Lower Control Limit of R chart = D3*4.15 (D3 = 0 for sample size 4) = 0 67
 Control Charts for Variable Type of Data
Example
Mean values and ranges of data from 20 samples (sample size = 4) are shown
in the table below:

Mean Mean Mean Mean Mean

S.N of Range S.N of Range S.N of Range S.N of Range S.N of Range
Sample Sample Sample Sample Sample

1 10 4 5 9 5 9 10 4 13 12 4 17 12 4

2 15 4 6 11 6 10 11 6 14 12 3 18 15 3

3 12 5 7 11 4 11 12 5 15 11 3 19 11 3

4 11 4 8 9 4 12 13 4 16 15 4 20 10 4
20

X
i 1
Average of mean values of 20 samples =
20
= 11.6 (Center Line of x bar Chart)

Upper Control Limit of x bar chart = 11.6 + A2*4.15 = 14.63 (A2 = 0.729 for sample size 4)

Lower Control Limit of x bar chart = 11.6 - A2*4.15 = 08.57 (A2 = 0.729 for sample size 4)
68
 Control Charts for Variable Type of Data
10

8
All the data are
within the LCL and
6
UCL in R Chart.
4
Hence variability of
2 the process data is not
0 an issue to worry.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

R Chart
Sample data at S.N 2,
16
16, and 18 are slightly
14
above the UCL.
12
Efforts must be made
10

8
to find the special
6
causes and revised
4
limits are advised to
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 calculate after
X-Bar Chart
deleting these data. 69
 Control Charts for Variable Type of Data
Consider a process by which coils are manufactured. Samples of size 5 are
randomly selected from the process, and the resistance values (in ohms) of the
coils are measured. The data values are given in Table

R i
87
R i 1
  3.48
g 25

For a sample of size 5, D4 = 2.114 and D3 = 0. The trial control limits for the
R-chart are calculated as follows:

UCLR  D4 R  (2.114)(3.48)  7.357

LCLR  D3 R  (0)(3.48)  0

70
Control Charts for Variable Type of Data
Sample Observation (ohms) X R Comments

1 20,22,21,23,22 21.60 3
2 19,18,22,20,20 19.80 4
New vendor
3 25,18,20,17,22 20.40 8
4 20,21,22,21,21 21.00 2
5 19,24,23,22,20 21.60 5
6 22,20,18,18,19 19.40 4
7 18,20,19,18,20 19.00 2
8 20,18,23,20,21 20.40 5
9 21,20,24,23,22 22.00 4
10 21,19,20,20,20 20.00 2
11 20,20,23,22,23 21.00 3
12 22,21,20,22,23 21.60 3
13 19,22,19,18,19 19.40 4
14 20,21,22,21,22 21.20 2
15 20,24,24,23,23 22.80 4
16 21,20,24,20,21 21.20 4
17 20,18,18,20,20 19.20 2
18 20,24,22,23,23 22.40 4
19 20,19,23,20,19 20.20 4
20 22,21,21,24,22 22.00 3
21 23,22,22,20,22 21.80 3
22 21,18,18,17,19 18.60 4 High Temperature
Wrong Die
23 21,24,24,23,23 23.00 3
24 20,22,21,21,20 20.80 2
25 19,20,21,21,22 20.60 3

Sum=521.00 Sum=87 71
Control Charts for Variable Type of Data
g

X i
521.00
X i 1
  20.840
g 25

UCL X  X  A 2 R  20.84  (0.577)(3.48)  22.848

LCL X  X  A 2 R  20.84  (0.577)(3.48)  18.832

72
Control Charts for Variable Type of Data

73
 Control Charts for Variable Type of Data
The revised control limits on the R-chart are
UCLR  D4 R  (2.114)(3.273)  6.919
LCLR  D3 R  (0)(3.273)  0

The revised center line on the X BAR chart

459
X  20.864
22

The revised control limits on the

UCL X  X  A 2 R  20.864  (0.577)(3.273)  22.753

LCL X  X  A 2 R  20.864  (0.577)(3.273)  18.975

Note that sample 15 falls slightly above the upper control limit on the chart.
On further investigation, no special causes could be identified for this sample.
So, the revised limits will be used for future observations until a subsequent
revision takes place. 74
 Control Charts for Variable Type of Data
Variable Sample Size
A change in the sample size has an impact on the control limits for the
X bar and R charts

An increase in the sample size n reduces the width of the control limits.

the width of the control limits from the center line is inversely
proportional to the square root of the sample size.

Standardized Control Charts

–When the sample size varies, the control limits on an X bar and an R-chart will
change, as discussed previously. With fluctuating control limits, the rules for
identifying out-of-control conditions become difficult to apply—that is, except for
Rule 1 (which assumes a process to be out of control when an observation
plots outside the control limits). One way to overcome this drawback is to use a
standardized control chart.

75
Control Charts for Variable Type of Data
g

n X i i
X  i 1
g
X i  sam ple m ean
n
i 1
i

^  ( n i  1) s i2
  i 1
g
s i  s a m p le S D
 i 1
( n i  1)

Xi  X
Zi  ^
/ ni
+3

0
Zi values are plotted on
a CC with CL=0
-3
UCL & LCL +3 & -3
76
Control Charts for Variable Type of Data
Ri
ri  ^

ri  d2
Ki 
d3

+3

0
Ki values are plotted on
a CC with CL=0
-3
UCL & LCL +3 & -3
77
 Control Charts for Variable Type of Data
Control limits for given target or standard

 given
X process mean (target)
0 st.dev
For R chart
CL X  X0 ^ R
3 0 
UCL X  X0   X 0  A 0 d2
n
^
3 0
LCL X  X0   X 0  A0  R   d2 0  given
n

CLR = d20
UCLR = D20
LCLR = D10
Care must be taken while interpreting CC based on standard values
-Target value is specified as too high or too low 78
Control Charts for Variable Type of Data
Refer to the coil resistance data, Let's suppose the target
values for the average resistance and standard deviation are
21.0 and 1.0 ohms, respectively. The sample size is 5. The
center line and the control limits for the

-Xchart are as follows:

CL X  X 0  21.0
UCL X  X 0  A  0  21.0  (1.342)(1.0)  22.342
LCL X  X 0  A  0  21.0  (1.342)(1.0)  19.658

The center line and control limits for the R-chart are
CLR = d20 = (2.326)(1.0) = 2.326
UCLR = D20 = (4.918)(1.0) = 4.918
LCLR = D10 = (0)(1.0) = 0

79
 Control Charts for Variable Type of Data

The process seems to be out of control with respect to the given standard.
Samples 5 and 8 are above the upper control limit, and a majority of the
points lie above the center line. Only six of the points plot below the center
line. thus it reveals that the process is not capable of meeting the company
guidelines Sigma as 1.0. The estimated process standard deviation from
(calculated after the process was brought to control) is
 R 3 .5 0
σ = = = 1 .5 0 5
d2 2 .3 2 6 80
Control Charts for Variable Type of Data

Several points are out-side the control limits—four points below

and two points above. the revised center line for the X bar chart
was found to be 20.864. Our target center line is now 21.0.
Adjusting controllable process parameters could possibly
shift the average level up to 21.0. However, the fact that
there are, points outside both the upper and lower control
limits signifies that process variability is the issue here. 81
 Control Charts for Variable Type of Data
and s charts

Centre line, upper, & lower control limit for s charts are calculated.
The formulae used are as following:

Suppose we have g preliminary samples at our disposition,

each of size n, and let si be the standard deviation of the ith sample.
Then the average of the g standard deviations is

g


1
S  si (Center Line for ‘s’ chart) Where B3 and B4
g i 1 are constants and
their values can
UCLR  B4 S (Upper control limit for ‘s’ chart) be obtained from
standard tables.
These values
LCLR  B3 S (Lower control limit for ‘s’ chart) depend upon 82the
size of sample.
Control Charts for Variable Type of Data
Factors for Calculating Limits for and S Charts

n A3 B3 B4
2 1.880 0 3.267
3 1.023 0 2.575
4 0.729 0 2.282
5 0.577 0 2.115
6 0.483 0 2.004
7 0.419 0.076 1.924
8 0.373 0.136 1.864
9 0.337 0.184 1.816
10 0.308 0.223 1.777

83
 Control Charts for Variable Type of Data
Centre line, upper, & lower control limit for x bar charts are calculated.
The formulae used are as following:
n

X i
Xi  i 1

X i = mean of the i sample

th

n = sample size,
Xi = ith data
g

X X = mean of ‘g’ samples

X  i 1
i
 CL
g x
Where A3 is
constants and its
value can be
obtained from
UCL x  X  A3 S (Upper control limit for X bar chart)
standard table. Its
value depend
LCL x  X  A3 S (Lower control limit for X bar chart) upon the size of
84
sample.
 Control Charts for Variable Type of Data
Example
Mean values and SD of data from 30 samples (sample size = 5) are shown in
the table below:

SN Mean SD SN Mean SD SN Mean SD SN Mean SD SN Mean SD

1 55.6 9.63 7 46.8 6.72 13 44 14.35 19 50.2 7.6 25 44.6 8.96
2 61 8.63 8 44.2 8.53 14 51.6 5.18 20 44 8.46 26 46.8 6.5
3 45.2 7.4 9 50.8 11.95 15 53.2 5.36 21 50 5.15 27 49.2 3.19
4 46.2 4.09 10 48.4 6.19 16 52.4 9.48 22 47 5.15 28 45.6 7.96
5 46.8 7.22 11 51.2 6.83 17 50.6 3.44 23 50.6 5.55 29 57.6 14.38
6 49.8 8.76 12 49.4 5.46 18 56 7 24 48.8 6.5 30 51.4 6.8

g


1
Average of Ranges of 30 samples = S  s i = 7.41 (Center Line of s Chart)
g i 1

Upper Control Limit of s chart = B4*7.41 (B4 = 2.115 for sample size 5)
= 15.48
Lower Control Limit of R chart = B3*7.41 (B3 = 0 for sample size 5) = 0
85
 Control Charts for Variable Type of Data
Example
Mean values and SD of data from 30 samples (sample size = 5) are shown in
the table below:

SN Mean SD SN Mean SD SN Mean SD SN Mean SD SN Mean SD

1 55.6 9.63 7 46.8 6.72 13 44 14.35 19 50.2 7.6 25 44.6 8.96
2 61 8.63 8 44.2 8.53 14 51.6 5.18 20 44 8.46 26 46.8 6.5
3 45.2 7.4 9 50.8 11.95 15 53.2 5.36 21 50 5.15 27 49.2 3.19
4 46.2 4.09 10 48.4 6.19 16 52.4 9.48 22 47 5.15 28 45.6 7.96
5 46.8 7.22 11 51.2 6.83 17 50.6 3.44 23 50.6 5.55 29 57.6 14.38
6 49.8 8.76 12 49.4 5.46 18 56 7 24 48.8 6.5 30 51.4 6.8
30

X
i 1
Average of mean values of 30 samples = = 49.63 (Center Line of x bar Chart)
30
Upper Control Limit of x bar chart = 49.63 + A3*7.41 = 60.21 (A3 = 1.427 for sample size 5)

Lower Control Limit of x bar chart = 49.63 - A3*7.41 = 39.05 (A3 = 1.427 for sample size 5)
86
 Control Charts for Variable Type of Data

60
Sample data at S.N 2, is
55 slightly above the UCL.
S am p le M ean

Efforts must be made to find

50 the special causes and
revised limits are advised to
45 calculate after deleting these
data.
40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

X Bar chart

14

10 All the data are within the

Sample SD

LCL and UCL in R Chart.

6 Hence variability of the
process data is not an issue
2
to worry.
-2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

S chart
87
Control Charts for Variable Type of Data
Example
Data for Magnetic Coating Thickness (in microns)
Sample Sample
Sample Mean Sample Standard
Sample Sample Mean Standard
X Deviation, s
X Deviation, s
1 36.4 4.6 11 36.7 5.3
2 35.8 3.7 12 35.2 3.5
3 37.3 5.2 13 38.8 4.7
4 33.9 4.3 14 39.0 5.6
5 37.8 4.4 15 35.5 5.0
6 36.1 3.9 16 37.1 4.1
7 38.6 5.0 17 38.3 5.6
8 39.4 6.1 18 39.2 4.8
9 34.4 4.1 19 36.8 4.7
10 39.5 5.8 20 37.7 5.4
88
Control Charts for Variable Type of Data
The control limits for the s-chart are
UCL s = B 4 s = (2.266)(4.790) = 10.854
LCL s = B3 s = (0)(4.790) = 0

89
 Control Charts for Variable Type of Data
The center line of the X Bar Chart is
20

 X i
7 4 3 .5
UCL X = X + A 3 s = 37.175 + (1.628)(4.790) = 44.973
CL X = X = i =1
= = 3 7 .1 7 5
20 20 LCL X = X - A 3 s = 37.175 - (1.628)(4.790) = 29.377

90
 Control Charts for Variable Type of Data
Assuming the thickness of the coating to be normally
distributed, what proportion of the product will not meet
specifications?
 s 4.790
z1 =
33.5-37.175
=-0.71 z =
42.5-37.175
σ= = = 5.199
5.199 2 =1.02
c 4 0.9213 5.199

Hence, the proportion of product not meeting specifications is 0.2389+0.1539=0.3928.

91
Control Charts for Variable Type of Data
Control limits for given target or standard

– Limits for s chart

CLs = C40
UCLs= B60
LCLs = B50

– Limits for X bar chart

UCL X = X0 + A 0
LCL X = X0 - A 0
92
Control Charts for Variable Type of Data
–Control Charts for Individual Units:
–100% inspection
–Destructive Testing
–Quality characteristic is represented as X,
–Variability as MR
σ =
MR
d2
C L MR = MR
 U C L M R = D 4 M R 
  n = 2 ,D 3 = 0 , D 4 = 3 .2 6 7
 L C L M R = D 4 M R 
CL X = X
 MR 
UCL = X + 3 
 d2 
  we can plot sp.lim it over control chart.
LCL = X - 3 MR  93
 d2 
 
 Control Charts for Variable Type of Data
Example
Table below shows the Brinell hardness number (BHN) of 20 individual
steel fasteners and the moving ranges. The testing process dents the parts so
that it cannot be used for their intended purpose. Construct the X- chart and
the MR- chart based on two successive observations. Specification limits
are 32±7.

Sample BHN MR Sample BHN MR

1 36.3 - 11 29.4 1.1

2 28.6 7.7 12 35.2 5.8
3 32.5 3.9 13 37.7 2.5
4 38.7 6.2 14 27.5 10.2
5 35.4 3.3 15 28.4 0.9
6 27.3 8.1 16 33.6 5.2
7 37.2 9.9 17 28.5 5.1
8 36.4 0.8 18 36.2 7.7
9 38.3 1.9 19 32.7 3.5
94
10 30.5 7.8 20 28.3 4.4
 Control Charts for Variable Type of Data
Example
σ = MR CLX = X
d2  MR 
CLMR = MR UCL= X+3 
 d2 
 wecanplot sp.limit over control chart.
UCLMR = D 4 MR  LCL = X-3MR 
  n = 2,D = 0,D = 3.267
LCLMR = D 4 MR 
3 4
 d2 


 
 MR 96   X i658 .7
  32 .90
MR    5.053
i X
20 20
19 19
 3 * 5 .053
UCL x  32 .9   46 .339
UCLMR  D4 MR  3.267* 5.053  16.058 1 .128
 3 * 5 .053
LCLMR  D3 MR  0 * 5.053  0 LCL x  32 .9   19 .461
1 .128

95
Control Charts for Variable Type of Data
Example

96
 Control Charts for Variable Type of Data
Cumulative Sum Control Chart for the Process Mean

X & R Chart → plotted point represents information for that objects only
Cumulative Sum Chart: CUSUM Chart
→uses information from all of the prior samples by displaying the
cumulative sum of the deviation of the samples values from a specified target
value. m

The cumulative sum at sample no. m is Sm =  (Xi - μ0 )

i=1

CUSUM chart are more effective than C.C in detecting relatively

small shift in the process mean. A cusum chart uses information from
previous samples, so that the effect of a small shift is more pronounced.

→ used for well established processes that have a history of being

stable.
To check the significance of shift, a template known as v-mask
97
(Barnard, 1959) is used
Control Charts for Variable Type of Data
Cumulative Sum Control Chart for the Process Mean




98
Control Charts for Variable Type of Data
Cumulative Sum Control Chart for the Process Mean

d &  Determination:
X
δ=
σ
X

δ  D egree of shift in the process m ean, relative to S D of m ean

 X  Am ount of shift in the process m ean that w e w ant to be
able to defect
σ  S .D of X
X

2  1- β 
d= ln   β << 0
δ2  α 
2
d=- ln(α)
δ 2

 X 
θ = tan -1  
 2k 
99
k is a scale factor (k should be σ & 2σ w ith a preferred value of 2 σ )
X X X
 Control Charts for Variable Type of Data
Example
Cumulative Sum Control Chart for the Process Mean

In the preparation of a drug, the percentage of calcium is a

characteristic we want to control. Random samples of size 5 are
selected, and the average percentage of calcium is found. The data
values from 15 samples are shown. From historical data, the standard
deviation of the percentage of calcium is estimated as 0.2%. The target
value for the average percentage of calcium content is 26.5%. We
decided to notice shifts in the average percentage of calcium content of
0.1%. Assume an acceptable Type I error level of .05.

100
 Control Charts for Variable Type of Data
Example
Cumulative Sum Control Chart for the Process Mean

Table: Average Percentage of Calcium

Sample Average Percentage Sample Average Percentage Sample Average Percentage

of Calcium, X of Calcium, X of Calcium, X
1 25.5 6 25.9 11 26.9
2 26.0 7 27.0 12 27.8
3 26.6 8 25.4 13 26.2
4 26.8 9 26.4 14 26.8
5 27.5 10 26.3 15 26.6

101
 Control Charts for Variable Type of Data
Example
Cumulative Sum Control Chart for the Process Mean

Cumulative Sum of Data for Calcium Content

Sample Deviation of sample Cumulative Sample Deviation of sample Cumulative

i Mean from Target Sum, Si i Mean from Target Sum, Si
(X- o) (X- o)
1 -1.0 -1.0 9 -0.1 -1.4
2 -0.5 -1.5 10 -0.2 -1.6
3 0.1 -1.4 11 0.4 -1.2
4 0.3 -1.1 12 1.3 0.1
5 1.0 -0.1 13 -0.3 -0.2
6 -0.6 -0.7 14 0.3 0.1
7 0.5 -0.2 15 0.1 0.2
8 -1.1 -1.3 102
 Control Charts for Variable Type of Data
Example

We must first find the deviation of each sample mean X, from the target mean o
= 26.5 and then fed the cumulative sum si. These values are shown in Previous
Table. From the information given, σ = 0.2, so X =  /n = 0.2/5 = 0.089. Next,

ΔX 0.1
δ= = = 1.124
σ 0.089
X
The lead distance for the V-mask is
2 2
d=- ln α = - ln(.05) = 4.472
δ 2
2 1.124
 ΔX 
θ = tan-1  
 2k 
The angle of the V-mask for K=0.125 is  0.1 
= tan-1  
 2(0.125) 
= tan-1(0.40) = 21.80o103
 Control Charts for Variable Type of Data
Example

This cumulative sum plot is shown. Using a V-mask with d = 4.742 and  =21.80o, a
downward shift in the process mean is detected rather quickly by the second sample.
When the V-mask is positioned on the cumulative sum point for sample 2, the cusum
for sample 1 is found to lie above the upper decision line of the V-mask, indicating
104
that the process has shifted downward.
 Control Charts for Variable Type of Data
Example
Moving-average control chart
Standard Shewhart control charts are quite insensitive to small shifts and
CUSUM charts are one way to alleviate this problem. A control chart
using the moving average is another way to alleviate the problem.
n  sample size
the means of first t samples are
X 1 , X 2 .......... X t 1 ,X t

w  moving average width ( w samples )

  
X t X t 1  ........ X t  w 1
M t 
w 105
 Control Charts for Variable Type of Data
Example
Moving-average control chart

 2
t 
Var (M t )  when w
n t

CL  X  X i
M  i  1
 t
t
UCL  X  3
nw CL  X

UCL  X  3 
nw UCL  X  3
nt

UCL  X  3
nt

106
 Control Charts for Variable Type of Data
Example
Moving-average control chart
The amount of a coloring pigment in polypropylene plastic,
produced in batches, is a variable of interest. For 20 random samples
of size 5, the average amount of pigment (in kilograms) is shown in
Table. Construct a moving-average control chart of width 6. The
process has up to this point been in control with an average range R
of 0.40 kg.
20
 Xt
503.2
X = i=1 = = 25.16
20 20
Since R = 0.40, an estimate of the process standard deviation is
 R 0.40
σ= = = 0.172
d 2.326
2 107
Control Charts for Variable Type of Data: Moving Av. Example
Moving Average Control Limits for Mt
Sample Sample Average, Xt
Mt LCL UCL
1 25.0 25.0 24.929 25.391
2 25.4 25.2 24.997 25.323
3 25.2 25.2 25.037 25.293
4 25.0 25.15 25.045 25.275
5 25.2 25.16 25.057 25.263
6 24.9 25.21 25.066 25.254
7 25.0 25.12 25.066 25.254
8 25.4 25.12 25.066 25.254
9 24.9 25.07 25.066 25.254
10 25.2 25.10 25.066 25.254
11 25.0 25.07 25.066 25.254
12 25.7 25.20 25.066 25.254
13 25.0 25.20 25.066 25.254
14 25.1 25.15 25.066 25.254
15 25.0 25.17 25.066 25.254
16 24.9 25.12 25.066 25.254
17 25.0 25.12 25.066 25.254
18 25.1 25.02 25.066 25.254
19 25.4 25.08 25.066 25.254 108
20 25.8 25.20 25.066 25.254
Control Charts for Variable Type of Data: Moving Av. Example

The moving average for sample 18 plots below the lower

control limit, indicating that the process mean has drifted
downward. Special causes should be investigated for this out-
of-control condition, and appropriate corrective action should
109
be taken.
Control Charts for Attribute Type of Data

Control chart for attributes

Attribute: - Quality characteristic for which a numerical
value is not specified.
Taste, paint quality, SF, Quality of output,
–Nonconformity (defect)
–Nonconforming (defective)
Attribute - at all levels, cost effective
Variable - at lowest level
 rules are not used because distribution in attributes
charts are non-normal

110
Control Charts for Attribute Type of Data
– Attribute information indicates whether a certain quality
characteristic is within specified limits. It does not state
the degree to which specifications are met or not met.
Thus Variable control chart provide more information on the
of performance of the process
– Attribute charts require larger sample sizes

N >> large
If p = 2.5%, n=25,
nonconforming item per
sample = 0.625
 p = 2.5%, n =100,
nonconforming item per
sample = 2.5
111
Control Charts for Attribute Type of Data

–(p-chart) Binomial Distribution

–for BD:- validity the Pnc=constant from sample to
sample and the sample must be independent.

n!
P( X  x)  p x (1  p ) n  x
(n  x)! x!
p  prob.of nonconforming
n  sample size
E(p̂)  p
p(1 - p)
var(p̂) 
n 112
Control Charts for Attribute Type of Data

p chart: chart for proportion nonconforming

p-charts calculates the percent defective in sample. p-charts

are used when observations can be placed in two categories
such as yes or no, good or bad, pass or fail etc.

113
 Control Charts for Attribute Type of Data
Centre line, upper, & lower control limit for p charts are calculated.
The formulae used are as following:

sum of defectives in all samples = centre line of p chart

p
total number of items in all samples

p (1  p ) (Upper control limit for p chart)

UCL  p  3
n

p (1  p)
LCL  p  3 (Lower control limit for p chart)
n
Where n is the sample size. Sample size in p chart must be  50

Sometimes LCL in p chart becomes negative, in such cases LCL should be taken114
as 0
 Control Charts for Attribute Type of Data
Example
Data for defective CDs from 20 samples (sample size = 100) are shown in the
table below:
Sample No. No. of Proportion Sample No. No. of Proportion
Defective Defective = Defective Defective =
CDs = x x/sample size CDs = x x/sample size
1 4 .04 11 6 .06
2 3 .03 12 5 .05
3 3 .03 13 4 .04
4 5 .05 14 5 .05
5 6 .06 15 4 .04
6 5 .05 16 7 .07
7 2 .02 17 6 .06
8 3 .03 18 8 .08
9 5 .05 19 6 .06
10 6 .06 20 8 .08

p (1  p ) . 051 (1  .051 )
UCL  p  3  0 .051  3  . 12
sum of defectives 101
n 100
CL =   0 . 051
sum of allsamples 2000
p (1  p ) .051(1  .051)
LCL  p  3  0 .051  3   .02  0
n 100 115
Control Charts for Attribute Type of Data
P CHART

0.14

0.12

0.10
MEANP
0.08 UCLP
0.06 LCLP
P
0.04

0.02

0.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

There is important observation that is clearly visible from the data points that
there is an increasing trend in the average proportion defectives beyond
sample number 15. Also, data show cyclic pattern. Process appears to be out
of control and also there is strong evidence that data are not from
independent source. 116
Control Charts for Attribute Type of Data

Variable Sample Size:

– 100% inspection, a change in the rate of production
may cause the sample size to change
– Change in the unit cost of inspection
– Change in the available inspection personnel
Control Limits for individual samples

p (1  p )
UCL  p  3
ni

p (1  p ) – Example 8.3 Amitava Mitra

LCL  p  3
ni 117
Control Charts for Attribute Type of Data
Control Limits based on Average sample Size

g
n  
i 1
ni

p (1  p )
UCL  p  3
n

p (1  p )
LCL  p  3
n

118
Control Charts for Attribute Type of Data

Special Considerations for p charts

– Necessary Assumption:
Probability of occurrence of nonconformity is constant
and Items are assumed to be independent of each other
– Observation below the control limit
– Comparison with a specified standard

119
Control Charts for Attribute Type of Data

np chart: chart for the number of nonconforming

np-charts calculates the number of defective in sample. np-

charts are used when observations can be placed in two
categories such as yes or no, good or bad, pass or fail etc.

It is easier to relate the number of defectives than to the

proportion defective. However, the np charts should not be
used when the sample size varies as the Centre line and the
control limits will vary.

120
 Control Charts for Attribute Type of Data
Centre line, upper, & lower control limit for np charts are calculated.
The formulae used are as following:

sum of defectives
np  = centre line of np chart
total number samples

 
   (Upper control limit for np chart)
UCL  n p  3 n p 1  p 
 

 
   (Lower control limit for np chart)
LCL  n p  3 n p 1  p 
 
Where n is the sample size. Sample size in np chart must be  50

Sometimes LCL in np chart becomes negative, in such cases LCL should be taken
121as 0
 Control Charts for Attribute Type of Data
Example
Data for the number of defective products from 20 samples (sample size =
300) are shown in the table below:
Sample No. No. of Sample No. No. of
Defective Defective
1 10 11 6
2 12 12 19
3 8 13 10
4 9 14 8
5 6 15 7
6 11 16 4
7 13 17 11
8 10 18 10
9 8 19 6
10 9 20 7

 
UCL  n p  3 n p (1  p )  9 . 2  3 9 . 2 (1  9 . 2 )  18 . 16
300
sum of defectives 184
CL =   9 .2
Number of samples 20  
LCL  n p  3 n p (1  p )  9 . 2  3 9 . 2 (1  9 . 2 )  0 . 24
300
122
Control Charts for Attribute Type of Data
20
18
16
14
12
Defectives

10
8
6
4
2
0
-2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Sample data at S.N 12, is above the UCL. Efforts must be made to find the
special causes and revised limits are advised to calculate after deleting these
data.

123
 Control Charts for Attribute Type of Data
–(c-chart) Poisson Distribution
–for PD
1. Opportunity area should be large
2. Occurrence of non-conformity must be independent to each other
3. Each sample should have equal chance of the occurrence of
nonconformity

e cc x
p( x) 
x!
p  prob.of x nonconform ing
x  number of nonconform ity in the sample
c  mean

124
Control Charts for Attribute Type of Data

c charts: Chart for the number of nonconformities

c-charts counts the number of defects in an item. c-charts are used only
when the number of occurrence per unit of measure can be counted such
as number of scratches, cracks etc.

125
 Control Charts for Attribute Type of Data
Centre line, upper, & lower control limit for c charts are calculated.
The formulae used are as following:

sum of defects in all samples

c = centre line of c chart
total number of items in all samples

(Upper control limit for c chart)

UCL  c  3 c

LCL  c  3 c (Lower control limit for c chart)

126
 Control Charts for Attribute Type of Data
Example
Data for defects on TV set from 20 samples (sample size = 10) are shown in
the table below:

Sample No. of Sample No. of Sample No. of Sample No. of

No. Defects No. Defects No. Defects No. Defects
1 5 6 4 11 6 16 5
2 4 7 5 12 5 17 4
3 5 8 6 13 4 18 6
4 6 9 8 14 7 19 6
5 4 10 7 15 6 20 6

sum of defects 109

CL =   5.45
numberof samples 20

UCL  c  3 c  5.45  3 5.45  12.45

LCL  c  3 c  5.45  3 5.45  1.55  0

127
Control Charts for Attribute Type of Data

14

12

10
Defects

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

None of the sample is out of the LCL and UCL. But the chart shows cyclic
trend.

128
Control Charts for Attribute Type of Data

u charts: Chart for number of nonconformity per unit

u-chart counts the number of defect per sample. The u chart is used when
it is not possible to have a sample size of a fixed size.

129
 Control Charts for Attribute Type of Data
Centre line, upper, & lower control limit for u charts are calculated.
The formulae used are as following:
k

sum of defects in all samples c i

u =
i 1
k = centre line of u chart
total number of items in all samples n
i 1
i

ci =number of defects in ith sample

k = number of samples
ni = size of ith samples

u
UCL  u  3 (Upper control limit for u chart)
ni

u (Lower control limit for uc chart)

LCL  u  3
ni 130
 Example
Data for defects in a product from 20 samples are shown in the table below.
The number of products varies from sample to sample. Construct a control
for number of defects per 100 products.

Sample No. of No. of Sample No. of No. of Sample No. of No. of

No. products Defects No. products Defects No. products Defects
inspected inspected inspected
1 200 5 8 150 10 15 100 6
2 300 14 9 150 6 16 200 8
3 250 8 10 250 10 17 200 5
4 150 8 11 300 9 18 100 5
5 250 12 12 250 16 19 300 14
6 100 6 13 200 12 20 200 8
7 200 20 14 250 10

4.68
c
k
UCL1  4.68  3  9.27

u i 1
i


192
 4.68
2
k
41
n i 4.68
i 1
LCL1  4.68  3  0.092 131
2
S No. Sample size No. of UCL LCL S No. Sample size No. of UCL LCL
ni Defects per ni Defects per
100 product 100 product
1 2.0 2.50 9.3 0.1 11 3.0 3.00 8.4 0.9
2 3.0 4.67 8.4 0.9 12 2.5 6.40 8.8 0.6
3 2.5 3.20 8.8 0.6 13 2.0 6.00 9.3 0.1
4 1.5 5.33 10.0 0.0 14 2.5 4.00 8.8 0.6
5 2.5 4.80 8.8 0.6 15 1.0 6.00 11.2 0.0
6 1.0 6.00 11.2 0.0 16 2.0 4.00 9.3 0.1
7 2.0 10.00 9.3 0.1 17 2.0 2.50 9.3 0.1
8 1.5 6.67 10.0 0.0 18 1.0 5.00 11.2 0.0
9 1.5 4.00 10.0 0.0 19 3.0 4.67 8.4 0.9
10 2.5 4.00 8.8 0.6 20 2.0 4.00 9.3 0.1

12

10

8
Defects

-2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

132
 CHART FOR DEMERITS PER UNIT (U-CHART)

• c and u charts do not differentiate with the types of

nonconformity
• There is a need to have a system that assigns weights to
nonconformities according to their relative degree of
severity.
• Charts for demerits per unit (U chart) takes care of the
deficiency of the c and u charts.

133
 CHART FOR DEMERITS PER UNIT (U-CHART)

Class 1 defects- very serious.

Defects that lead directly to sever injury or to catastrophic economic loss.

Class 2 defects- serious.

Defects that lead to significant injury or significant economic loss.

Class 3 defects- major.

Defect that can cause major problems with normal use of a product or service
rendered.
Class 4 defects- minor
Defect that can cause minor problems with normal use of a product or service
rendered.

134
 CHART FOR DEMERITS PER UNIT (U-CHART)

For the ANSI/ASQC standard, a weight system of 100, 50, 10 and 1

could, for example, be chosen for the categories of very serious,

serious, major and minor, respectively.

135
 CHART FOR DEMERITS PER UNIT (U-CHART)

Suppose we have four categories of nonconformities. Let the

sample size be ‘n’ and let c1, c2, c3 and c4 denote the total number
of nonconformities in a sample for four categories. Let w1, w2,
w3, and w4 denote the weights assigned to each category.
It is assumed that nonconformities in each category are
independent of defects in the other categories. Furthermore, it is
also assumed that the occurrence on nonconformities in any
category is represented by a Poisson distribution.

136
CHART FOR DEMERITS PER UNIT (U-CHART)

For a sample of size ‘n’, the total number of demerits is given

by

D = w1c1 + w2c2 + w3c3+ w4c4 (1)

The demerits per unit for the sample are given by

(2)

137
 CHART FOR DEMERITS PER UNIT (U-CHART)

The center line of the U-chart is given by

(3)

Where represent the average number of

nonconformities per unit in their respective classes.

138
 CHART FOR DEMERITS PER UNIT (U-CHART)

The estimated standard deviation of U is given by

(4)

The control limits for the U‐chart are given by

(5)

If the lower control limit is calculated to be less than zero, it is converted to zero.
139
 CHART FOR DEMERITS PER UNIT (U-CHART)

Example: A department store obtains feedback on customer

satisfaction regarding a certain product. Twenty random samples,
each involving 10 customers, are taken in which customers are
asked about the number of serious, major, and minor
nonconformities that they have experienced. Clear definitions of
each category are provided. The weights assigned to a serious,
major, and minor nonconformity are 50, 10, and 1, respectively.
Construct a control chart for the number of demerits per unit.

140
 Table: Data for nonconformities in a department store customer survey
Serious Major Minor Total Demerits per
Sample Nonconformities Nonconformities Nonconformities Demerits unit
c1 c2 c3 D U
1 1 4 2 92 9.2
2 0 3 8 38 3.8
3 0 5 10 60 6.0
4 1 2 5 75 7.5
5 0 6 2 62 6.2
6 0 0 8 8 0.8
7 0 7 5 75 7.5
8 1 1 1 61 6.1
9 1 3 2 82 8.2
10 0 4 12 52 5.2
11 1 5 3 103 10.3
12 2 0 2 102 10.2
13 0 0 9 9 0.9
14 0 6 8 68 6.8
15 1 12 10 180 18.0
16 0 5 7 57 5.7
17 0 1 1 11 1.1
18 1 2 5 75 7.5
19 0 5 6 56 5.6
20 0 3 8 38 3.8
141
Total 9 74 114
 CHART FOR DEMERITS PER UNIT (U-CHART)

Solution:
For each sample, the total number of demerits given by equation (1) is
shown in Table. The table also shows the number of demerits per unit U,
given by equation (2). To find the center line Ū, the average number of
nonconformities per unit for each category is calculated.
For (Serious” nonconformities

Similarly,

74
u2   0.37
200
114
u3   0.57
200 142
 CHART FOR DEMERITS PER UNIT (U-CHART)

Using eq. (3), the center line of the U-chart is

The estimated standard deviation of U, using eq. (4) is

Hence, the control limits (from eq. (5)) are

L 143
 CHART FOR DEMERITS PER UNIT (U-CHART)
UCL= 18.142

CL= 6.52

LCL= 0

Figure U-chart for department store customer survey.

Note: Figure shows (U‐chart) all the point within the control limits 144